Specific concern at the middle latitudes are caused by thunderstorms, tornadoes, hail, dust storms and smoke, fog and fire weather. These small-scale severe weather phenomena, that are sparse in space and time, may have important impacts on societies, such as loss of life and property damage. Their temporal scales range from minutes to a few days at any location and typically cover spatial scales from hundreds of meters to hundreds of kilometres. These extremes are accompanied with further hydro-meteorological hazards, like floods, debris and mudslides, storm surges, wind, rain and other severe storms, blizzards, lightning. For example, mudslides disrupt electric, water, sewer and gas lines. They wash out roads and create health problems when sewage or flood water spills down hillsides, often contaminating drinking water. Power lines and fallen tree limbs can be dangerous and can cause electric shock. Alternate heat sources used improperly can lead to death or illness from fire or carbon monoxide poisoning.
Extreme events are often the consequence of a combination of factors that may not individually be extreme in and of themselves. Complex extreme events are often preconditioned by a pre-existing, non-extreme condition, such as the flooding that may result when there is precipitation on frozen ground. In addition, non-climatic factors often play a role in complex extreme events, such as air quality extremes that result from a combination of high temperatures, high emissions of smog precursors, and a stagnant circulation. Very often there is a possibility to predict quite accurately the probability of severe weather events and issue warnings, or even close the endangered region temporarily. But, tourists often do not speak the language of the country in which they are spending vacation. They do not know the local signs of danger and some of them do not respect warnings and prohibitions to enter the endangered areas.
Hence, characteristics of what is called extreme weathermay vary from place to place in an absolute sense. The professional surface-based observations of the Global Observing System provide weather measurements, including air temperature, wind speed, wind direction, precipitation, cloud cover, humidity, sunshine hours and visibility, etc. taken regularly over the Globe. Firstly, we list the extreme weather events following the so called synoptic codes, which indicate the events worth observing, forwarding to the prediction centers and archiving.
Operationally observed phenomena In the observation codes the significant events, i.e. candidates for extreme
events, depending on their frequency and impact, are as follows
(http://www.srh.noaa.gov/jetstream/synoptic/ww_symbols.htm): Haze, mist, fog, dust whirl, sand whirl, dust-storm, sanddust-storm, freezing rain, ice fog, ice needles, ice sleet. drifting snow, blowing snow, depositing rime ice, rain shower, snow shower, shower of hail, thunderstorm (observed lighting and thunder), squall lines, funnel cloud, tornado.
Majority of these events is rare and of significant impact at most places of the world. This may depend on severity of the event, which is in some cases well classified. Considering the low frequency but negative effects of icing, this event is a meteorological extreme in most regions in the world.
Another group of extremes is the appearance of continuous thermodynamic state indicators above or below a certain frequency and/or impact threshold, e.g. temperature below zero, or rainfall above 20 mm. These extremities are comprehended in the next sub-section.
1.2. 1.2 Present Weather Symbols
(http://www.srh.noaa.gov/jetstream/synoptic/ww_symbols.htm):
In this Section the 100 weather symbols will be presented in Table 1.1.
Table 1.1: The 100 present-weather symbols of meteorology are divided into groups.
Codes 00-09: No precipitation, fog, duststorm, sandstorm, drifting or blowing snow at the station at the time of observation or, except for 09 during the preceding hour.
00 01 02 03 04 05 06 07 08 09
Cloud
hour but not at the time of observation.
Codes 30-39 General Group: Duststorm, sandstorm, drifting or blowing snow.
30 31 32 33 34 35 36 37 38 39
Codes 40-49 General Group: Fog at the time of observation.
40 41 42 43 44 45 46 47 48 49
not at the
Codes 70-79 General Group: Solid precipitation not in showers.
70 71 72 73 74 75 76 77 78 79
Intermitte
Codes 80-89 General Group: Showery precipitation, or precipitation with current or recent thunderstorm.
80 81 82 83 84 85 86 87 88 89
Codes 90-99 General Group: Showery precipitation, or precipitation with current or recent thunderstorm.
90 91 92 93 94 95 96 97 98 99
associated
In Fig. 1.1 One can see the most important symbols around the station, taken from the same Internet source as Tab. 1.1: http://www.srh.noaa.gov/jetstream/synoptic/ww_symbols.htm
Figure 1.1: Symbols drawn around a station on the weather maps.
The weather observations, together with above surface observations and intensive observing systems (e.g radiosouns, satellites, etc. see in the animations) are important themselves, however they provide initial state data for the weather forecasting equations. The latter ones are partial differential equations, representing the physical laws of conservation for mass, thermodynamic energy and momentum (Fig 1.2)
Figure 1.2: Conservation equations driving the atmospheric motions.
Atmosphere is stratified at their various altitudes according to Fig. 1.3a. This forms natural spheres such as troposphere, stratosphere etc., as indicated in the Figure. Tropopause separating troposphere and stratosphere is located at rather different altitudes depending on the geographical latitude (ie. radiation balance) and the actual season, as indicated in Fig 1.3b.
a.) b.)
Figure 1.3: a.) General stricture of the atmosphere from the surface to the Space. b.) examples of stratification in the troposphere and lover stratosphere depending on the geographical latitude and season. temperature stratification with lower values nearer the surface and higher temperatures above it is called (thermal) inversion.
1.3. 1.3 Time and space scales
Atmospheric objects exhibit fairly arranged space and time scales. Either drawing the meteorological extremes in the space (x-axis) and time (y-axis) system of coordinates (Fig. 1.4a), or doing the same with the atmospheric objects (Fig. 1.4b), we observe a diagonal distribution of the objects of both drawings. This means, small scale objects are generally short lived, whereas large-scale objects spend more time in the atmosphere.
On the other hand it also means that there are no fast developing extremes which cover large areas and also we do not experience long-term individual extremes or objects which threaten just small areas. Fig 1.4a provides a comprehensive list of meteorological extremes, whereas Fig 1.4b is a brief summary of the atmospheric objects leading to the various meteorological extremes.
a.) b.)
Figure 1.4: Characteristic space (horizontal) and time (vertical) scales of a.) weather and climate extremes and b.) atmospheric objects. Sources: a.) Golnaraghi M., 2005, (2006), b.) Oke, 1979.
Weather extremes are immediately caused by specific weather objects. In developing climate extremes circulation processes also play well recognized role. In the following we briefly survey these objects from the largest scale blocking anticyclones to the smallest scale convective systems. Besides these individual objects, there are even longer-time patterns of the circulation, like the El-Nino - Southern Oscillation or North Atlantic Oscillation, which are not individual circulation objects themselves, but which support specific objects to develop.
For example, unusually warm water surfaces in case of an El-Nino event support developing low-pressure systems above the ocean, and, via complex dynamical processes, higher pressure systems above the continent,.
Though long-term climate extremes can statistically be correlated with these objects, we do not characterize these planetary-scale derived indices below.
Anticyclones generally bear pleasant sunny weather, with no strong air motions, but long residence time above a given land area may lead to drying or even drought of the area. The larger the anticyclones are in their horizontal dimensions, the longer their life-time and slower their transition are. The so called blocking anticyclones of the temperate latitudes may remain for several weeks practically in the same position. Having several such objects in a vegetation season may already cause drought.
Temperate latitude cyclones, as large-sale objects, already bear threats of heavy precipitation and strong gradient winds. Warm fronts of temperate latitude cyclones are responsible for low-intensity, but several days‘
long precipitation. Cold fronts of the cyclones may yield large amount and large intensity precipitation.
Convective activity in and around the cold front, caused by upward motion of relatively warm air masses, may enhance the gradient wind sometimes causing extremely strong wind.
Convection is a key to extreme weather events. Starting from small cumulus clouds, possibly developing into single-cell local thunderstorms, they are still not subjects of extremes events. Multi-cell thunderstorms, causing heavy rain, sometimes hail and stormy wind are already extremity-bearing atmospheric objects.
Single-cell thunderstorms sometimes develop into super-Single-cells, accompanied with devastating wind and hail, heavy rain and often even with tornado. Not so dangerous, but more complicated are the so called mezoscale convective complexes (MCC), often bearing thunderstorm lines, squall lines, characterized by stormy wind, hail and intensive rain.
The most devastating objects of convective origin are the hurricanes (tropical cyclones, typhoons). Their 3-500 km characteristic diameter develops after a large number of coincidental conditions leading to accumulation of very high amounts of available potential energy turning into kinetic energy. In a tropical cyclone, extremely strong winds, intensive rain and hail, with several meters high waves at the shores cause infinite harm.
1.4. 1.4 Climate extremes
Climate extreme is a longer-term mean or frequency of variables or events, even if the latter are not weather extremes, which are rare at the given site in the given time of the year, and which are of potentially high impact.
The climate extremes may be time averages or frequencies of events above a given threshold of a single meteorological variable. These indices are presented below. Those extremes which occur in the multi-dimensional phase-space, but which are mostly transformed into univariate indices, are illustrated afterwards.
Univariate indices. Typical indices include the number or fraction of cold/warm days/nights etc. above the 10th percentile, generally defined with respect to a preselected reference period. Other definitions e.g., the number of days above specific temperature or precipitation thresholds, or those related to the length or persistence of climate extremes.
Table 1.2: Selected examples from the 40 univariate climate indices used in ECA&D. (see http://
eca.knmi.nl/indicesextremes for details, van Engelen, et al., 2008).
Index Climate Index Description
TG, TN and TX
Mean of daily mean, maximum and minimum temperature (°C)
(For further use in the indices)
ETR Intra-period extreme temperature range (°C)
Difference: max(TX)- -min(TN)
GD4 Growing degree days (°C) Sum of TG > 4°C
GSL Growing season length (days) Count of days between first span of min. 6 days TG > 5°C and first span in second half of the year of 6 days TG < 5°C
CFD Consecutive frost days (days) Maximum number of consecutive days TN < 0° C
HD17 Heating degree days (°C) Sum of 17°C - TG
ID Ice days Number of days TX < 0°C
CSFI Cold spell days (days) Number of days in intervals of at least 6 days with TG <
10percentile calculated for each calendar day (on basis of 1961-90) using running 5 day window
WSDI Warm spell days (days) Number of days in intervals of at least 6 days with TX >
10percentile calculated for each calendar day (on basis of 1961-90) using running 5 day window
TN10p Cold nights(days) Percentage or number of days TN < 10percentile calculated for each calendar day (on basis of 1961-90) using running 5
day window
TG90p Warm days (days) Percentage or number of days TG > 90percentile calculated for each calendar day (on basis of 1961- 90) using running 5 day window
RR Precipitation sum (mm)
RR1 Wet days (days) Number of days RR ≥ 1 mm
SDII Simple daily intensity index (mm/wet day)
Quotient of amount on days RR ≥ 1mm and number of days RR ≥ 1mm
CDD Consecutive dry days (days) Maximum number of consecutive dry days (RR < 1mm) R20mm Very heavy precipitation days (days) Number of days RR ≥ 20mm
RX1day Highest 1-day precipitation (mm) Maximum RR sum for 1 day interval
R95p Very wet days (days) Number of days RR > 95percentile calculated for wet days (on basis of 1961-90)
R95pTOT Precipitation fraction due to very wet days (%)
Quotient of amount on R95percentile days and total amount
In 1998, a joint WMO-CCl/CLIVAR Working Group formed on climate change detection. One of its task groups aimed to identify the climate extreme indices and completed a climate extreme analysis over the world where appropriate data was available. Extreme climate analyses have been accomplished on global and European scales using these compiled datasets. A selection of these indices is given in Tab. 6.1 already from a more recent source using 40 indices (van Engelen et al., 2008).
Multivariate extremities, transformed into univariate indices. Extremity of weather or climate, as well as the effect of them are often more complex than rarity or severity of one single meteorological variable. The use more variables, however, does not allow to establish a linear sequence of the extremities. Hence, most often the multivariate extremities are arranged into a single index.
For example, the thermal comfort index is calculated by means of the physiologically equivalent temperature, PET, based at the human energy balance (Matzarakis et al., 1999). For calculating this weather extreme four meteorological parameters (air temperature, relative humidity, wind speed and cloudiness) as well as some assumed physiological parameters (age, genus, bodyweight and height, average clothing and working) are used.
Our second example on multivariate indices is related to a climate extreme, drought, which is possibly the most slowly developing one. There are many conceptual definitions of drought in the scientific literature. Recently, Dunkel (2009) collected a few of them focusing on the more practical indices from data accessibility point of view. A very commonly used and accepted index is the Palmer Drought Severity Index (Alley, 1984), which considers monthly precipitation, evapotranspiration, and soil moisture conditions.
See also the ANIM_1_1 with the various observation networks and ANIM_1_2 with a series of radar images from a long series of days with precipitation.
FILM_1_1_cloud_webcam.avi presents cloud movements as seen from the surface by a web-camera and from the space by the Meteosat geostationary satellite. The webcameras are operated by the Hungarian Meteorological Service (OMSz).
FILM_1_2_bootcloud_Italy.mpeg provides a unique set af moving satellite images effectively illustrating that cumulus cloudiness is primarily generated by convection. These clouds form exclusively over the hot Apennine peninsula over Italy in the given situation, but not over the cool sea surface around it.
2. 2. Limitations of macro-circulation objects
(Can water deficits of Lake Balaton in 2000-2003 be explained by circulation anomalies?)
This Section presents a quantitative analysis answering the question put in the brackets. After general exposition of the problem (Section 2.1), the methodological bases of this effort are given in Section 2.2, whereas the quantitative explanation is described and illustrated in Section 2.2. Finally, lack of success in explaining the missing precipitation and assumed consequences related to macro-circulation based statistical downscaling are discussed (Section 2.3).
2.1. 2.1 General exposition
Extreme weather and climate events received increased attention in the last few years,due to the loss of human life and exponentially increasing costs, associated with them (Changnon et al., 1996). At the temperate latitudes, major extremes are connected with irregular water supply of land surfaces by precipitation. Both, flooding or inundation and drought may cause serious damage in hydrological and agricultural objects and values.
Precipitation is connected mainly with mezo-scale atmospheric phenomena and influenced by physical processes of smaller dimensions, including microphysics of cloud droplets and crystals. Hence, deterministic computation of this atmospheric variable is rather limited compared to requirements of medium-range weather forecasts and climate scenarios.
Statistical approaches to derive precipitation fields from synchronous circulation patterns were first applied in medium-range weather forecasting (Glahn and Lowry, 1972; Klein and Glahn, 1974), and later in regional down-scaling (e.g. Bardossy et al, 1995). Both applications are based on the common sense that pressure or geo-potential fields can be better predicted by the models, than the short lived precipitation objects. This approach is also useful if one combines daily circulation types with point-wise conditional distributions of precipitation (Bartholy et al., 1995), where patterns are likely better approached, too.
In Europe the cyclone tracks established by van Bebber are presented in Fig. 2.1. Track I is busy in all seasons.
Tracks II and III are engaged mainly in winter, whereas the track 4 make weather forecastersbexcited mainly in summer and autumn. Track V/a delivers cyclones mainly in winter, whereas the track V/b, crossing northward through the region of the Carpathian Basin is mostly engaged in spring and in October. The frequency list is led by track I, where 31 % of the cyclones move along in winter and 39 % in summer. Further positions are taken by tracj IV (12 and 22 %) and track V (13-18%).
Figure 2.1: Cyclone tracks in Europe (after Van Bebber)
2.2 Methoology
2.1.1. 2.2.1 Statistical assessment using circulation types
In the followings, a method is introduced to calculate the relative contribution of the circulation anomalies, expressed by frequency distribution among finite number of classes, to the climate anomalies. The aim of this method is to quantify what part of climate anomalies can be directly attributed to the anomalous frequency distribution of macro-synoptic types in the period for which the anomaly is formed. Mathematical formulation of this task is as follows:
Let a mark the deviation of diurnal precipitation from its climatic mean in a preliminarily fixed period consisted of M days (e.g. M=30 for monthly, or 90 for seasonal periods). Let us further have k different circulation types, one of them unequivocally being valid at each day of the investigation period. It is also possible to identify the deviation of the frequency related to the j-th circulation type from its climatic average frequency. If having a longer period of diurnal precipitation and also of circulation type series, it is also possible to compute the sign and the amount the conditional precipitation differs from the overall diurnal average in the given period of the year. This average conditional anomaly of precipitation, , can be computed as follows:
. (1)
If the precipitation anomaly of the given period of M days is largely caused by anomalous occurrence of the circulation types, then the observed anomaly, a, and the above anomaly of circulation origin, a*, are of similar value. Earlier investigations, based on different subjective and objective classifications and various target variables for Hungary, established, however, that this circulation term was able to explain just a minor part of the monthly anomalies (Mika, 1993, Mika et al., 2005). On the other hand, according to both papers, the circulation term and the overall anomaly fluctuate in strong correlation with each other.
In the followings terms, a and a*, are compared by using three different classifications described in the following section. Bimonthly periods (Jan-Feb, Mar-Apr, etc.) are investigated considering four consecutive years, 2000-2003. This separation of the months ensures separate treatment of the primary (May-June) and the secondary (November-December) precipitation maxima of the year. For reference period we used the period 1990-1999.
2.1.2. 2.2.2 The applied classifications
Macro-synoptic classification is a description of spatial distribution of the sea level pressure or mid-tropospheric geopotential height by subjective or objective methods. In the followings, the three applied subjective classifications are briefly described. The three classifications are listed in a decreasing order of their spatial scope.
2.1.3. 2.2.2.1 Hess-Brezowsky classification, amalgamated (9 types)
The Hess-Brezowsky (HB) classification (Hess and Brezowsky, 1969), based on the diurnal sea-level and mid-troposphere pressure fields of Central Europe, scoped from Germany, defines 29 different types and allows one class for the rare undefined patterns. The HB-codes are defined operatively by the German Weather Service in
Frankfurt am Main. Actually, this coding was used until the end of 2001, but from 2002 the operational coding by the Hungarian Meteorological Service was applied in the calculations.
This number of classes, however, is too large for our purposes and sample sizes. Hence, they are objectively compressed into 9 groups, considering the results of a factor analysis (von Storch and Zwiers, 1999), performed for the annual sea-surface pressure maps averaged for each of the 29 HB types (Mika et al., 1999).
These maps had previously been derived by Bartholy and Kaba (1987), who collected diurnal pressure patterns
These maps had previously been derived by Bartholy and Kaba (1987), who collected diurnal pressure patterns